(Publisher of Peer Reviewed Open Access Journals)

International Journal of Advanced Technology and Engineering Exploration (IJATEE)

ISSN (Print):2394-5443    ISSN (Online):2394-7454
Volume-8 Issue-80 July-2021
Full-Text PDF
Paper Title : Haar-features training parameters analysis in boosting based machine learning for improved face detection
Author Name : Vikram Mutneja and Satvir Singh
Abstract :

Haar features have been used in most of the works in literature as key components in the task of object as well as face detection. Training of Haar features is an important step in the development of overall machine learning based face detection system. In this work, we have done investigation in the variations of a number of training parameters during AdaBoost based machine learning of Haar features with respect to the size of training images. A number of observations have been drawn based on the variations of these parameters during the training process. Training parameters, true detection rate and figure of merit have been used as weighing parameter. These parameters have been used in the formation of detection cascade for the improvement in the original AdaBoost framework used for machine learning of Haar features. Statistical analysis based on variations of training parameters has been done for selection of efficient learners from a large pool of available features which further are cascaded to make the strong classifiers. We have been able to achieve the maximum percentage reduction in training time of 47.20 corresponding to training image’s size of 25×25 pixels with the help of statistical screening of Haar features prior to AdaBoost. The proposed system has been developed using training dataset from Center for Biological & Computational Learning (CBCL) and successfully tested on facial images from datasets. The datasets considered were WIDER FACE Detection Benchmark, Annotated Faces in the Wild (AFW) and Face Detection Dataset and Benchmark (FDDB). The results achieved were promising in terms of training time computed in the range of 0.31-2.83hrs. The maximum detection rate is 96.21% and minimum detection time is 16.95 ms.

Keywords : Haar features training, AdaBoost, Machine learning, Statistical analysis, Face detection, Features reduction.
Cite this article : Mutneja V, Singh S. Haar-features training parameters analysis in boosting based machine learning for improved face detection. International Journal of Advanced Technology and Engineering Exploration. 2021; 8(80):919-931. DOI:10.19101/IJATEE.2021.874076.
References :
[1]Viola P, Jones M. Rapid object detection using a boosted cascade of simple features. In proceedings of the 2001 computer society conference on computer vision and pattern recognition 2001. IEEE.
[Crossref] [Google Scholar]
[2]Viola P, Jones MJ. Robust real-time face detection. International Journal of Computer Vision. 2004; 57:137-54.
[Crossref] [Google Scholar]
[3]Egorov AD, Shtanko AN, Minin PE. Selection of viola–jones algorithm parameters for specific conditions. Bulletin of the Lebedev Physics Institute. 2015; 42:244-8.
[Crossref] [Google Scholar]
[4]Mena AP, Mayoral MB, Diaz-lope E. Comparative study of the features used by algorithms based on viola and jones face detection algorithm. In international work-conference on the interplay between natural and artificial computation 2015 (pp. 175-83). Springer, Cham.
[Crossref] [Google Scholar]
[5]Pandey S, Sharma S. An optimistic approach for implementing viola jones face detection algorithm in database system and in real time. International Journal of Engineering Research and Technology. 2015; 4(7):1118-22.
[Google Scholar]
[6]Mutneja V, Singh S. Modified Viola–Jones algorithm with GPU accelerated training and parallelized skin color filtering-based face detection. Journal of Real-Time Image Processing. 2019; 16:1573-93.
[Crossref] [Google Scholar]
[7]Rusia MK, Singh DK, Ansari MA. Human face identification using lbp and haar-like features for real time attendance monitoring. In international conference on image information processing 2019 (pp. 612-6). IEEE.
[Crossref] [Google Scholar]
[8]Papageorgiou CP, Oren M, Poggio T. A general framework for object detection. In sixth international conference on computer vision 1998 (pp. 555-62). IEEE.
[Crossref] [Google Scholar]
[9]Lienhart R, Maydt J. An extended set of haar-like features for rapid object detection. In proceedings of the international conference on image processing 2002. IEEE.
[Crossref] [Google Scholar]
[10]Saha S, Démoulin V. ALOHA: an efficient binary descriptor based on Haar features. In international conference on image processing 2012 (pp. 2345-8). IEEE.
[Crossref] [Google Scholar]
[11]Nirmalkar S, Janghel R. Haar features selection using statistics in boosting based object detection. 2012.
[Google Scholar]
[12]Zhang S, Bauckhage C, Cremers AB. Informed Haar-like features improve pedestrian detection. In proceedings of the IEEE conference on computer vision and pattern recognition 2014 (pp. 947-54).
[Google Scholar]
[13]Wang YQ. An analysis of the Viola-Jones face detection algorithm. Image Processing On Line. 2014; 4:128-48.
[Crossref] [Google Scholar]
[14]Murphy TM, Broussard R, Schultz R, Rakvic R, Ngo H. Face detection with a Viola–Jones based hybrid network. IET Biometrics. 2017; 6(3):200-10.
[Google Scholar]
[15]Qezavati H, Majidi B, Manzuri MT. Partially covered face detection in presence of headscarf for surveillance applications. In international conference on pattern recognition and image analysis 2019 (pp. 195-9). IEEE.
[Crossref] [Google Scholar]
[16]Chung BW, Park KY, Hwang SY. A fast and efficient Haar-like feature selection algorithm for object detection. The Journal of Korean Institute of Communications and Information Sciences. 2013; 38(6):486-91.
[Crossref] [Google Scholar]
[17]Mutneja V, Singh S. Size-based performance analysis of haar-features for detection of facial images from low resolution surveillance videos. In international conference on intelligent computing, instrumentation and control technologies 2017 (pp. 1191-5). IEEE.
[Crossref] [Google Scholar]
[18]Jin J, Zhu A, Wang Y, Wright J. A feature binding model in computer vision for object detection. Multimedia Tools and Applications. 2021; 80:19377-97.
[Crossref] [Google Scholar]
[19]Lin YN, Hsieh TY, Huang JJ, Yang CY, Shen VR, Bui HH. Fast Iris localization using haar-like features and AdaBoost algorithm. Multimedia Tools and Applications. 2020; 79:34339-62.
[Crossref] [Google Scholar]
[20]Adeshina SO, Ibrahim H, Teoh SS, Hoo SC. Custom face classification model for classroom using haar-like and LBP features with their performance comparisons. Electronics. 2021; 10(2):1-15.
[Crossref] [Google Scholar]
[21]Pant DR, Sthapit R. Analysis of micro facial expression by machine and deep learning methods: haar, CNN, and RNN. Journal of the Institute of Engineering. 2021; 16(1):95-101.
[Crossref] [Google Scholar]
[22]Mordvintseva YA, Belitsky AA. Training the haar cascade classifier in sorting radio-electronic components. automating the collection of images for teaching. In conference of Russian young researchers in electrical and electronic engineering 2021 (pp. 1019-22). IEEE.
[Crossref] [Google Scholar]
[23]Odaudu SN, Adedokun EA, Salaudeen AT, Marshall FF, Ibrahim Y, Ikpe DE. Sequential feature selection using hybridized differential evolution algorithm and Haar cascade for object detection framework. Covenant Journal of Informatics and Communication Technology. 2020; 8(1):67-82.
[Google Scholar]
[24]Wang YG, Li M, Ma Z, Montufar G, Zhuang X, Fan Y. Haar graph pooling. In international conference on machine learning 2020 (pp. 9952-62). PMLR.
[Google Scholar]
[25]Wu H, Cao Y, Wei H, Tian Z. Face recognition based on haar like and euclidean distance. In journal of physics: conference series 2021 (pp. 1-7). IOP Publishing.
[Crossref] [Google Scholar]
[26]Yang S, Luo P, Loy CC, Tang X. Wider face: a face detection benchmark. In proceedings of the conference on computer vision and pattern recognition 2016 (pp. 5525-33).IEEE.
[Google Scholar]
[27]Zhu X, Ramanan D. Face detection, pose estimation, and landmark localization in the wild. In conference on computer vision and pattern recognition 2012 (pp. 2879-86). IEEE.
[Crossref] [Google Scholar]
[28]Jain V, Learned-miller E. Fddb: a benchmark for face detection in unconstrained settings. UMass Amherst Technical Report; 2010:1-11.
[Google Scholar]
[29]http://cbcl.mit.edu/software-datasets/FaceData2.html. Accessed 26 March 2015.
[30]Fuhl W, Eivazi S, Hosp B, Eivazi A, Rosenstiel W, Kasneci E. BORE: boosted-oriented edge optimization for robust, real time remote pupil center detection. In proceedings of the ACM symposium on eye tracking research & applications 2018 (pp. 1-5).
[Crossref] [Google Scholar]
[31]Fuhl W, Gao H, Kasneci E. Tiny convolution, decision tree, and binary neuronal networks for robust and real time pupil outline estimation. In ACM symposium on eye tracking research and applications 2020 (pp. 1-5).
[Crossref] [Google Scholar]
[32]Fuhl W. 1000 pupil segmentations in a second using haar like features and statistical learning. arXiv preprint arXiv:2102.01921. 2021.
[Google Scholar]
[33]Naido S, Porle RR. Face detection using colour and haar features for indoor surveillance. In 2nd international conference on artificial intelligence in engineering and technology 2020 (pp. 1-5). IEEE.
[Crossref] [Google Scholar]
[34]Hou Y, Wang M. Face detection method based on improved convolutional neural network. In international workshop on electronic communication and artificial intelligence 2020 (pp. 102-5). IEEE.
[Crossref] [Google Scholar]
[35]Naveen P, Sivakumar P. Pose and head orientation invariant face detection based on optimised aggregate channel feature. Annals of the Romanian Society for Cell Biology. 2021; 25(5):4368-90.
[Google Scholar]